Line-Frame Combined Multi-Slit Hyperspectral Imaging System for Simplified and High-Precision UAV Remote Sensing Solution

This paper presents a UAV-based hyperspectral imaging system that can simultaneously capture a frame image and three hyperspectral images, providing an efficient and precise approach to hyperspectral data acquisition. By utilizing the frame image to preserve relative pose information across consecutive frames, the system simplifies geometric correction and eliminates the need for high-precision directional geo-referencing devices, ground control points, or gimbal systems. At the same time, the system employs a multi-slit time-sharing sampling strategy to enable repeated spectral measurements while reducing sampling loss and maintaining data authenticity. A matched geometric correction algorithm, which combines single-slit image mosaicking and multi-slit fusion, is proposed to ensure precise spatial alignment even under challenging conditions. Furthermore, a robust radiometric calibration framework, incorporating cross-calibration and real-time corrections via a downwelling radiance measurement device, mitigates errors caused by variable solar illumination. These features enable reliable reflectance measurements and high-quality hyperspectral images. A rigorous evaluation framework, including both ground-based and UAV-based experiments, is conducted to comprehensively assess system performance. The results demonstrate superior sampling efficiency, improved geometric alignment accuracy, and greater radiometric consistency compared to conventional methods. The proposed system offers a cost-effective, lightweight, and versatile solution for hyperspectral imaging applications in agriculture, environmental monitoring, target exploration, and other fields, effectively addressing key limitations of existing technologies.